How to use file names as metadata when adding books

[user743]

How can i add books to my calibre library, with the file names saved as "series namber"

the file names are just numbers, which i want to be able to sort my books by later.

[eschwartz]

Under Preferences ==> Adding Books you can "read the metadata from the filename instead of the file contents" using the appropriate regex (it will explain details) BUT you cannot also get the embedded metadata. So the ONLY metadata you will have is that number.

[user743]

is there any way around that problem?!

maybe bulk editing the metadata later and using "<title>([^<]+)</title>" as the new title, if possible?

[eschwartz]

The problem is that you can only scan the book for metadata on import, and you can choose either embedded metadata or filename metadata but not both.

If you preserve the filename by using it as the metadata, you cannot later extract metadata from the file.

What I would like calibre to do is offer an option that filename metadata should override embedded metadata, or at least act as an optional supplementary source of lower priority.

Either way would allow both of us to save the original filename as meaningful custom metadata.

Kovid, would it be possible to add an option that filename can be used only to supply metadata that isn't already embedded in the file?

[kovidgoyal]

Turn off reading metadata from files
Add the file
Turn on reading metadata from files
Read metadata from the file using the edit metadata dialog (top right area)

This is a little tedious, but I personally dont have the motivation to make it less tedious, as I have never come across a book that has bad title/author metadata and good other kinds of metadata, with the exception of covers, which you can read from files, in bulk with the bulk edit metadata dialog.

[eschwartz]

Thanks for the info.

In case it motivates you to fix it, I'd like to explain why I'd use it. I keep a custom column with the original filename of my Amazon purchases which I use as my save template. This allows me to simply overwrite the old file with my calibre copy that has corrected metadata, and preserves the annotations for the book. Additionally when I delete a book from the device UI it leaves the annotations files behind allowing me to simply re-add the book from calibre and have my old annotations.

For this I rely on a stable save filename. It is easier if this is the filename Amazon uses as well, since it works with both Amazon and non-Amazon books and I can redownload the annotations from Amazon if necessary then just update the book by overwriting.

So I'd appreciate an enhanced add option -- but I understand if it's not worth the time to implement.

calibre is still awesome, and I have it all working smoothly albeit with a little more effort than my laziness requests of me.

[user743]

can you tell me the regex to save the filename in to a costom column called "filename".

edit
apparently that's not a option so how do i save it as series.
the file names look like this: with different 5 digit and "part" numbers.

f_01234
f_01567_all
f_01567_part_1
f_01567_part_23

thanks kovldgoyil for calibre and all your help!

[BetterRed]

Originally Posted by user743

can you tell me the regex to save the filename in to a costom column called "filename".

edit
how do i save it as series. the file names look like this: with different 5 digit and "part" numbers.

f_01234
f_01567_all
f_01567_part_1
f_01567_part_23

@user743 - In Preferences->Add your own columns you could try creating a custom column with type Text column for keeping series-like information

BR

[user743]

how do i add them two begin with, not in the title.

no matter what regex i use the whole filename becomes the title.

[eschwartz]

You can save to any column I believe...

Just ese a regex to read the filename and save the right section with a named capture group, named with the lookup name of the custom/column.